Compiling RenumKicadPCB/Linux Mint

User Jano has found that the compilation instructions for RenumKicadPCB under Linux Mint should include -lm (see https://documenteddesigns.com/2017/03/27/renumkicadpcb-v0203/)

“I just compiled and used your utility successful! Thanks. Just a note to compile I needed to ad -lm at the end of the gcc order: gcc RenumKiCadPCB.c -o RenumKiCadPCB -lm my system is linuxmint 18.3, after that it worked perfectly.”

My understanding is that -lm usually includes a library module (i.e. -lmMingw32) and I’m not a gcc guru so I don’t know what the -lm by itself does, but if you have problems give it a try.

Thanks Jano!

Hand Assembled SMD Board

I figured I’d post a photo of a board I’m working on.

It is an I2C to 20×4 LCD /keyboard  interface with support for an ESP8266-01 module plus the serial interface modules.

I ordered the solder paste screen from Elcrow and figured I’d give it a try. A couple years ago I picked up a convection toaster over from Walmart and modified it so the fan would run always on.

There are lots of tutorials and such about making a complex controller for “toaster oven reflow” but my approach is simple: I figured out the setting on the temperature dial so the solder melts then I made two marks on the dial: that point and a bit cooler.

I “pre-heat” the board at the cooler setting then move the dial to the “melt” setting for about 5 minutes, then move back to “pre-heat” after the solder melts, then after a couple minutes and shut the unit off.

This seems to work fine. My view is that the fixation with temperature ramps and the like is because those parameters are for volume production and that makes a huge difference if you are making a thousand units an hour.

This board was manually assembled using my technique. I had a bit of problem with the solder paste, probably because I used a silicone spatula (for cooking) rather than a proper spreader. The solder paste was about 3 years old but kept in the refrigerator.

DSCN0118

The picture was taken straight out of the oven, after I removed a bridge across Q1. The assembly is not perfect but that is mainly because of hand assembly and solder paste application, not the oven. In particular, I’d be happier if U4 was aligned properly but the soldering seems OK.

The board was designed and laid out with Kicad.

RenumKicadPCB V0203

I wasn’t expecting to post a new version so soon but while preparing some features I discovered an actual bug! Here it is RenumKicadPCBV0203

The bug showed itself when sorting on reference designations and the component orientation was not typical. Unfortunately the Kicad file formats are not very well documented and I did not realize that the reference designation offset was relative to the module’s angular rotation, not absolute. Therefore if the ref des offset was 0.1, 0.2, that doesn’t mean it was 0.1 to the right and 0.2 down relative to the module – it means it was 0.1, 0.2 relative to the original grid of the component. Its actual location is a trigonometry problem which is simple enough to figure out once you figure out what is going on.

So, the adjustment is

X offset = (( Ref_X_Offset * cos( modangle )) + ( Ref_Y_Offset * sin(modangle )));

Y offset = (( -Ref_X_Offset* sin(modangle )) + ( Ref_Y_Offset * cos(modangle )));

In addition to this bug fix, I removed the “Verbose” and “Show Change Array” menu options and instead create three log files:

  1. _coord.txt, which shows the module or reference designation coordinates and their “rounded to grid function;
  2. _update.txt which shows the nets, reference designations, etc., which will be renamed, and to what
  3. _change.txt which is just the reference designation change plan and show how modules were renamed by the code.

These files are pretty small and making them does not seem to have a significant impact on run time. They make it a lot easier to figure out the decisions the software made and its easier to ask for a file than to ask someone to “cut and paste” from a DOS terminal screen.

Besides the obscure bug corrected in the V0203, which I discovered, I am not aware of any errors made in sorting, file updating, etc.. Thus far unexpected results are not the fault of the software but a mismatch between the options used and the design. I actually found the bug because my “test project” ice-zum, has the reference designation for C1 in the place of C2 and vice versa so when sort on modules it showed C2 C1. It was only while correcting that design mistake I realized there was an unrelated bug.

I also cleaned up the code “map to grid” is done only once.

As always feel free to contact me if you have any issues, can’t get it to work right, etc..

 

RenumKicadPCB V0.202

Note: RenumKiCadPCB has been updated several times to enhance usability and fix bugs. Please check documenteddesigns.com for the most recent version.

 

This RenumKiCadPCBV0202 is the updated Windows 64 application and source code for V0.202 of RenumKiCadPCB.

I moved sorting from the confusing “jog” implementation to a grid based implementation. All module and ref designation locations are rounded to the nearest specified grid point (X, Y) prior to sorting. So a module at 23.1235, 19.568 would be sorted as 23.25, 19.50 if the grid is specified as 0.25.

Note that this does not actually move modules or text: RenumKiCadPCB only changes the names of things not their positions.

Also the “Sort on” default has been changed to “Sort on Modules” based on feedback I received from Nick Barton of Black Magic Amplifiers.

Although I usually sort by Reference Designations, Nick – who is a pro – believes “Sort on Modules” is a far more typical use case.

As always, you can change the “Sort on” parameter from the menu and since V0.201 these choices are “sticky” and saved for the next use.

Note that, unless reference designations, module coordinates are somewhat arbitrary and this may lead to strange results. For example if a diode has the cathode as the origin, two diodes side by side with one rotated (see below)

diodeexample

would be seen as separated by 1 cm (1/2″) on the Y axis whereas the reference designation text would be “on grid” and properly dealt with.

Dealing with “Bad” Renumbering

While there may be bugs in the code (please contact me directly if there are) most reports of “bad” renumbering are due to two reasons:

  1. Sorting of reference designations instead of modules or modules instead of reference designations; and
  2. Sorting of too fine a grid (or jog in older versions).

RenumKiCad PCB approaches the problem mathematically and if you have placed parts or reference designations 1 mm apart on the sort axis and the sort grid (or jog) is less than 1 mm it will see the parts as being on different lines and number them accordingly.

Remember – the software doesn’t know if you have done a coarse through hole design on a 2.54 mm (0.1″) grid or an ultra-fine pitch layout with 0.1 mm (0.005″) grid, nor does it know whether you have your module origins all in the “right place”.

The options to sort on modules or reference designation and to change the grid are there to provide flexibility.

So if you get a “bad” sort order, try changing the sort target from modules to reference designations and/or adjust the sort grid.

Usually increasing the sort grid fixes things up to a point as making it too large has the opposite effect as all parts will appear to be at the same point.

So try increasing it to 0.25, 0.5, or 1 until you get the desired result. Because the sort works on the original data you can simply run RenumKiCadPCB over and over again on your design until you get what you want.

Always make backups of the original.

I don’t know of any bugs which would ruin your project but I am not responsible – or even want to feel bad – if it does.

The only way I can improve this software is if you provide feedback. Please let me know if you use it and, especially if there are any issues.

 

RenumKicadPCB V0.201

Note: RenumKiCadPCB has been updated several times to enhance usability and fix bugs. Please check documenteddesigns.com for the most recent version.

I have added a feature to RenumKicadPCB. The program now stores parameters in a file called Renumparameters.txt in the directory it is invoked from. This means you don’t have to keep reentering the path to the file, etc..

Here is a zip file with the instructions for use, source code, and a windows 64 executable renumkicadpcbv0-201

RenumKiCadPCB V0.200

Note: RenumKiCadPCB has been updated several times to enhance usability and fix bugs. Please check documenteddesigns.com for the most recent version.

 

This file RenumKiCadPCBv0200 is V0.200 of RenumKiCadPCB.

RenumKiCadPCB processes a KiCad PCB file and renumbers all the component reference designators ending in numbers based on where they are located on the PCB. It then processes the schematic hierarchy and updates the component reference designators to match. This makes working on a board much easier since you can locate all the components.

V0. 200 includes a number of improvements including a rudimentary text based menu driven interface as well as the original command line. Some command line arguments have been changed to make more sense and some defaults have been changed as well.

V0.200 was also developed on MinGW instead of Cygwin so there is no need for a separate DLL – the executable is stand alone.

To run to the program simply type RenumKiCadPCB (command line arguments are now optional).

I have made and tested the file on Windows (executable included) and Ubuntu Linux. It should work on Apple but I have not tested it.

I hope to port the code over to a proper GUI once I figure out how to set up the compiler, etc..

Let me know if there are any issues, bugs, feature requests.

 

 

V0.103 RenumberKiCadPCB

Note: RenumKiCadPCB has been updated several times to enhance usability and fix bugs. Please check documenteddesigns.com for the most recent version.

This RenumKiCadPCB.V0103 is an updated version of RenumberKiCadPCB. Earlier today I posted v0.102 and 0.103 updates net names to correct some (but not all) issues associated with netlist imports and pours.

Since I take a second pass on the file to update nets it takes about 100mSecs longer.

Thanks to Andy_P at https://forum.kicad.info/ who tested the program and correctly identified that it did not handle hierarchical sheets.

The program has been extensively rewritten to support hierarchies and (I believe) “channels” or the same sheet repeated with different reference designators.

I also added the option to -c output a change list and leave the files alone and a -y option so you don’t have to type Y/N to run the program.

Thanks also to PCB_Wiz on https://forum.kicad.info/ who graciously proved a link to this project https://github.com/FPGAwars/icezum/tree/master/src-kicad (which is a beautiful project by the way).

After running V0.101 on PCB_Wiz’s icezum project, it seems to work fine. There is an issue if I then create a schematic netlist from the updated files and import that into PCBNew and run the DRC. One netlist is flagged no connects even though they appear to be connected. The name of the problem net is neither in the original (pre-run) PCB file or the post run file.

I lack the KiCad skills to determine if this is a problem with my code or the way the original files were made (I have had KiCad complain about files it alone had made)

It appears to be easy enough to correct but the project is too complex for me to understand. Note that this only occurs if you recreate the netlist.

If anybody has a similar problem and can reproduce it on a minimal project I can probably fix the code if there is a bug.

Please let me know if there are any issues.